New MCP Work tools enable approved AI agents to read, write, search and progress legal work within existing permissions, workflows and audit controls
NASHVILLE, Tenn. – 24 August 2026 – Autologyx today announced that it is showcasing its latest MCP-enabled capabilities at ILTACON 2026, demonstrating how law firms can give AI agents practical access to legal work while maintaining control over what they can see, what they can do and how their actions are governed.
The new capabilities build on the MCP support announced by Autologyx earlier this year, which enables approved AI agents to participate directly in enterprise workflows rather than operating as standalone assistants outside the business process.
At ILTACON, Autologyx is demonstrating more than 35 MCP Work tools that allow approved agents to interact with work running through the platform, including the ability to read and update records, search data, create and complete tasks, generate documents and help progress matters through governed workflows.
The tools are designed to address a practical challenge emerging as law firms adopt more AI platforms and agentic capabilities: connecting agents to systems is becoming easier, but controlling what happens once those connections exist is significantly harder.
“Law firms are moving quickly from asking what AI can do to asking how they actually put it to work,” said Ben Stoneham, Founder and Chief Technology Officer at Autologyx.
“MCP gives firms a much more consistent way to connect agents to the systems and data they need. The next challenge is making sure those agents operate within the same client, matter, process and permission boundaries as everyone else.”
Autologyx applies its existing governance and orchestration architecture to agent activity, allowing firms to control access at a granular level based on factors including user, group, client, matter, process, dataset and workflow stage.
This means access to an MCP tool does not automatically give an agent unrestricted authority to use it.
Instead, firms can determine which agents and models are permitted to perform particular actions, against which data, and under what circumstances. Human approvals, business rules and workflow controls can remain part of the process where required.
The approach reflects the increasingly multi-vendor nature of legal AI.
Rather than requiring firms to standardise on a single AI provider, Autologyx is designed to provide a governed operational layer through which approved agents from different AI environments can participate in legal work.
“The reality is that firms are going to have multiple AI tools, models and agents,” Stoneham continued. “The governance problem therefore cannot live inside one AI product. Firms need an operational layer that understands the work itself and can apply the right controls regardless of which agent is making the request.”
Governed execution, not isolated AI
Autologyx sees MCP as an important part of the shift from AI assistance towards AI execution.
While many legal AI use cases to date have focused on research, drafting, summarisation and individual productivity, agentic systems are increasingly capable of taking actions inside business processes.
That might include updating a matter record, creating a task, retrieving information from a system of record, generating a document or triggering the next step in a workflow.
For law firms, those actions need to happen within existing operational and client controls.
Autologyx enables AI activity to sit alongside human tasks, deterministic automation, approvals and other system actions within the same workflow, providing firms with a single operational view of how work is being progressed.
Auditability across human and AI work
The latest capabilities also extend Autologyx's focus on auditability.
Agent activity can be captured alongside the wider workflow record, helping firms understand which agent or model performed an action, what information was involved, which tool was used and what happened as a result.
This provides a clearer record of AI-assisted work than treating AI activity as a separate layer outside the process.
The June launch of Autologyx's MCP capabilities introduced enhanced monitoring and audit functionality specifically to support this requirement, with visibility across model interactions, tool usage, workflow activity and resulting actions.
See governed MCP execution at ILTACON
Autologyx is demonstrating its MCP capabilities throughout ILTACON 2026 in Nashville.
Visitors can see how approved AI agents can work with live processes, data and systems through Autologyx while remaining subject to the permissions, workflow rules and audit controls already governing the work.
Autologyx can be found at Booth 836.
Run MCP. Keep control.
About Autologyx
Autologyx provides a governance and orchestration platform for complex legal and regulated work.
The platform brings together workflow orchestration, automation, relational data, integrations and applied AI, enabling organisations to coordinate human, system and AI activity within controlled and auditable processes.
Autologyx helps law firms and other regulated organisations connect AI to real work while maintaining visibility, accountability and control.